[Image quality in multidetector CT of paranasal sinuses: potential of dose reduction using an adaptive post-processing filter]

Purpose: Evaluation of subjective image quality in dose-reduced multi-detector CT (MDCT) of paranasal sinuses using a 2D non-linear adaptive post-processing filter (2D-NLAF).

Materials and methods: MDCT of paranasal sinuses was simulated using a human head phantom at a Somatom Sensation Cardiac 64 (Siemens, Erlangen). At constant collimation (64 × 0.6 mm) und pitch (p = 1), the tube current (50, 100, 200 mAs) and tube potential (80, 100, 120 kVp) were modified. The radiation exposure was represented by CTDIvol. Four independent blinded radiologists evaluated the image quality of axial 2 mm images and coronal reformations concerning the assessment of "fractures" and "soft tissue processes". The subjective image quality of original and post-processed images using a 2D-NLAF (SharpViewCT®, Sweden) was graded on a 5-point scale ("1" excellent - "5" not adequate) and compared.

Results: Compared to the protocol with the best image quality (120kVp/ 200 mAs) 2D-NLAF led to a significant improvement in the subjective image quality at 100 kVp/ 100 mAs (score "1.4" with filter versus "2.2" without) and 120 kVp/ 50 mAs ("1.6" versus "2.0") (p < 0.03) particularly for high contrasts ("fractures", p < 0.001). In "soft tissue processes", 2D-NLAF provided improved quality from "2.1" to "1.4" (p < 0.04) at 100 kVp/ 100 mAs. Down to a CTDIvol of 8 mGy, the image quality was rated "good", and down to 5 mGy "diagnostic".

Conclusion: The phantom study indicates a dose reduction potential in MDCT of paranasal sinuses up to 58% compared to a standard dose protocol using a 2D-NLAF without an essential loss of image quality. 2D-NLAF is particularly effective at 100 kVp/ 100 mAs and 120 kVp/ 50 mAs.